Linear Algebra and SVD (Some slides adapted from Octavia Camps)

Linear Algebra and SVD(Some slides adapted from Octavia Camps)

Goals

• Represent points as column vectors.

• Represent motion as matrices.

• Move geometric objects with matrix multiplication.

• Introduce SVD

Euclidean transformations

2D Translation

tt

PP

P’P’

2D Translation Equation

PP

xx

yy

ttxx

ttyy

P’P’tt

tPP ),(' yx tytx

),(

),(

yx tt

yx

t

P

2D Translation using Matrices

PP

xx

yy

ttxx

ttyy

P’P’tt

),(

),(

yx tt

yx

t

P

1

1

0

0

1' y

x

t

t

ty

tx

y

x

y

xP

tt PP

Scaling

PP

P’P’

Scaling Equation

PP

xx

yy

s.xs.x

P’P’s.ys.y

),('

),(

sysx

yx

P

P

PP s'

y

x

s

s

sy

sx

0

0'P

SPSP '

Rotation

PP

PP’’

Rotation Equations

Counter-clockwise rotation by an angle Counter-clockwise rotation by an angle

y

x

y

x

cossin

sincos

'

'

PP

xx

Y’Y’PP’’

X’X’

yy R.PP'

Why does multiplying points by R rotate them?

• Think of the rows of R as a new coordinate system. Taking inner products of each points with these expresses that point in that coordinate system.

• This means rows of R must be orthonormal vectors (orthogonal unit vectors).

• Think of what happens to the points (1,0) and (0,1). They go to (cos theta, -sin theta), and (sin theta, cos theta). They remain orthonormal, and rotate clockwise by theta.

• Any other point, (a,b) can be thought of as a(1,0) + b(0,1). R(a(1,0)+b(0,1) = Ra(1,0) + Ra(0,1) = aR(1,0) + bR(0,1). So it’s in the same position relative to the rotated coordinates that it was in before rotation relative to the x, y coordinates. That is, it’s rotated.

Degrees of Freedom

R is 2x2 R is 2x2 4 elements4 elements

BUT! There is only 1 degree of freedom: BUT! There is only 1 degree of freedom:

1)det(

R

IRRRR TT

The 4 elements must satisfy the following constraints:The 4 elements must satisfy the following constraints:

y

x

y

x

cossin

sincos

'

'

Transformations can be composed

• Matrix multiplication is associative.

• Combine series of transformations into one matrix. (example, whiteboard).

• In general, the order matters. (example, whiteboard).

• 2D Rotations can be interchanged. Why?

Rotation and Translation

cos -sin tx

sin cos ty

0 0 1

( )(x

y

1)

a -b tx

b a ty

0 0 1

( )(x

y

1)

Rotation, Scaling and Translation

Rotation about an arbitrary point

• Can translate to origin, rotate, translate back. (example, whiteboard).

• This is also rotation with one translation.– Intuitively, amount of rotation is same

either way.– But a translation is added.

Stretching Equation

PP

xx

yy

SSxx.x.x

P’P’SSyy.y.y

y

xs

s

ys

xs

y

x

y

x

0

0'P

),('

),(

ysxs

yx

yx

P

P

S

PSP '

Stretching = tilting and projecting(with weak perspective)

y

xs

s

sy

xs

s

ys

xsy

x

yy

x

y

x

10

00

0'P

SVD

• For any matrix, M = R1DR2

– R1, R2 are rotation matrices

– D is a diagonal matrix.– This decomposition is unique.– Efficient algorithms can compute this (in

matlab, svd).

Linear Transformation

y

xs

s

s

y

xs

s

y

x

dc

ba

y

x

y

y

x

sincos

cossin

10

0

sincos

cossin

sincos

cossin0

0

sincos

cossin

'PSVD

Affine Transformation

1

' y

x

tydc

txbaP

Simple 3D Rotation

n

n

n

zzz

yyy

xxx

21

21

21...

100

0cossin

0sincos

Rotation about z axis.

Rotates x,y coordinates. Leaves z coordinates fixed.

Full 3D Rotation

cossin0

sincos0

001

cos0sin

010

sin0cos

100

0cossin

0sincos

R

• Any rotation can be expressed as combination of three rotations about three axes.

100

010

001TRR

• Rows (and columns) of R are orthonormal vectors.

• R has determinant 1 (not -1).

• Intuitively, it makes sense that 3D rotations can be expressed as 3 separate rotations about fixed axes. Rotations have 3 degrees of freedom; two describe an axis of rotation, and one the amount.

• Rotations preserve the length of a vector, and the angle between two vectors. Therefore, (1,0,0), (0,1,0), (0,0,1) must be orthonormal after rotation. After rotation, they are the three columns of R. So these columns must be orthonormal vectors for R to be a rotation. Similarly, if they are orthonormal vectors (with determinant 1) R will have the effect of rotating (1,0,0), (0,1,0), (0,0,1). Same reasoning as 2D tells us all other points rotate too.

• Note if R has determinant -1, then R is a rotation plus a reflection.

3D Rotation + Translation

• Just like 2D case